WalkCast — Screen-Free Audio Digest for Outdoor Walks

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 1: Touch Grass

WalkCast is a 100% offline, screen-minimizing audio web app that ingests long-form articles, tech news, and newsletters, converting them into a warm, conversational podcast script.

Instead of staring at a smartphone screen while walking outside, WalkCast turns reading lists into an audio digest. Users can hit generate, put their phone in their pocket, keep their eyes on the road, and catch up on content during a 20-minute daily walk.

Because WalkCast is designed as a privacy-first, offline tool powered by local Llama 3.2 inference, it runs locally on the user's browser without a hosted backend.

html
<!-- Simply open walkcast.html in any browser with local Ollama running -->
## Code
// Fetch conversational podcast script from local Llama 3.2 instance
const response = await fetch('http://localhost:11434/api/generate', {
    method: 'POST',
    headers: { 'Content-Type': 'application/json' },
    body: JSON.stringify({
        model: 'llama3.2',
        prompt: `Convert this article into a warm, conversational 1-minute podcast summary for a walk: ${articleText}`,
        stream: false,
        options: { num_predict: 150 }
    })
});

// Play back using native browser SpeechSynthesis at walking speed (0.95x)
const utterance = new SpeechSynthesisUtterance(scriptText);
utterance.rate = 0.95;
window.speechSynthesis.speak(utterance);

## How I Built It
Local Inference: Meta's Llama 3.2 (3B) model running locally via Ollama.

Audio Synthesis: Browser-native SpeechSynthesis API tuned to 0.95x pacing to match a relaxing walking cadence.

Neighborhood UI: Designed with a cozy, storybook street aesthetic matching the theme of stepping outside and touching grass.
## Why Does Open Innovation Matter?
Screen Minimization & Health: Shifts content consumption from visual screens to audio, reducing eye strain and increasing pedestrian awareness while walking outdoors.

Complete Privacy: Personal reading lists and newsletters never leave your machine or get sent to third-party ad networks.

Zero Cost & Offline: Works without cloud API subscriptions or cellular data reliance while out on trails or neighborhood routes.